languageR (version 1.0)

acf.fnc: Autocorrelation trellis graph

Description

This function creates a trellis plot with autocorrelation functions for by-subject sequential dependencies in response latencies.

Usage

acf.fnc(dat, group="Subject", time="Trial", x = "RT", plot=TRUE, ...)

Arguments

dat

A data frame with (minimally) a grouping factor, an index for successive trails/events, and a behavioral measure

group

A grouping factor such as Subject

time

A sequential time measure such as Trial number in the experimental list

x

The dependent variable, usually a chronometric measure such as RT

plot

If true, a trellis graph is produced, otherwise a data frame with the data on which the trellis graph is based is returned

...

other optional arguments, such as layout

Value

• If plot=TRUE, a trellis graph, otherwise a data frame with as column names
• Lag
  Autocorrelation lag
• Acf
  Autocorrelation
• Subject
  The grouping factor, typically Subject
• ci
  The (approximate) 95% confidence interval.

References

R. H. Baayen (2001) Word Frequency Distributions, Dordrecht: Kluwer.

Examples

data(beginningReaders)
acf.fnc(beginningReaders, x="LogRT")   # autocorrelations even though nonword responses not included